Department of Economics
Working Paper Series
The Value of Adoption
by
Mary Eschelbach Hansen
No. 2006-15
December 2006
http://www.american.edu/cas/econ/workpap.htm
Copyright © 2006 by Mary Eschelbach Hansen. All rights reserved. Readers may make verbatim
copies of this document for non-commercial purposes by any means, provided that this copyright
notice appears on all such copies.
The Value of Adoption*
Mary Eschelbach Hansen**
Abstract
The human services cost of adoption is about half the cost of long-term foster care for children
whose birth parents’ rights have been terminated. Because adoption is an effective intervention for
improving a variety of outcomes for those exposed to adverse childhood experiences, the total
savings to government in areas such as special education and criminal justice is of the same
magnitude as the child welfare savings. The private benefit to adopted children in terms of
additional income earned over their working lives is similarly large. In all, a dollar spent on the
adoption of a child from foster care yields about three dollars in benefits.
Key words: adoption, adoption policy, cost-benefit analysis, foster care, net social
benefits
JEL classification: D61, J13, J18
*
This research was supported by a grant from the Center for Adoption Research at the University of
Massachusetts Medical School. The research assistance of Jodi Adams, Wendy Guyker, and Laura
Sherbin is much appreciated. Comments from Debra Chaison-Demers, Bradley Hansen, Jennifer E.
Macomber, Kara Reynolds, Audrey Smolkin, and John Willoughby have strengthened this paper, as
have conversations with Richard Barth and Rita Simon. Remaining errors are the property of the
author.
**
Department of Economics, American University. Contact: Mary Eschelbach Hansen, Department
of Economics, American University, 4400 Massachusetts Avenue, NW, Washington, DC 200168029. Email: mhansen@american.edu.
The Value of Adoption
I. Introduction
The total number of minors ever adopted after foster care approaches the size of the foster
care caseload (Wulczyn and Hislop 2002). It follows that state and federal expenditures for
adoption assistance under Title IV-E of the Social Security Act have grown. Expenditures grew
from less than $400,000 in fiscal year 1981 to $1.3 billion in fiscal year 2002 (Dalberth et al 2005).
The federal adoption assistance budget grew 30 percent between 2000 and 2002 alone (Scarcella et al
2004).
Since 2000, fiscal stress has led several states to attempt to cut adoption assistance spending
(North American Council on Adoptable Children [NACAC], 2003; Eckholm, 2005). Some of the
cuts have been blocked by the courts, which have made it clear that adoptive parents have legal
standing to protect their children’s entitlements (E.C. v. Blunt (05-0726-CV-W-SOW) and A.S.W. v.
Oregon (also known as A.S.W. v. Mink, 424 F. 3d 970 (9th Cir. 2005)). These decisions effectively
require states to take adoption assistance and adoption policy more seriously.
Cutting back on adoption spending may make sense if the cost of adoption exceeds its
benefits. Although it is generally believed that benefits to adoption are relatively high (see Barth
1997, for example), neither the private nor social benefit of adoption of children from foster care
has been estimated.
The benefits of adoption accrue to the adopted and to society at large. The psychic benefit
to the adopted of having a permanent family is, of course, inestimable. It is possible, though, to
estimate the private monetary value of adoption in terms of the higher lifetime incomes earned by
the adopted relative to those raised in long-term foster care. Offsetting the benefits of adoption are
private costs borne by the adoptive parents. It is also possible to estimate the benefits and costs of
1
adoption to government in terms of the streams of savings from adoption and expenditures on the
adopted.
An adoption from foster care costs state and federal government about $115,000, but saves
the government about $258,000 in child welfare and human service costs, netting a savings of
$143,000 (Barth et al 2006, adjusted for inflation to 2000 dollars). I show that each adoption nets
between $88,000 and $150,000 in private benefits and $190,000 to $235,000 in total public benefits
(in constant 2000 dollars). Thus each dollar spent on the adoption of a child from foster care yields
between $2.45 and $3.26 in benefits to society.
II. Method
In this paper I employ traditional methods of cost-benefit analysis. I limit the scope of the
study to the private and social benefits and costs of adoption relative to long term foster care over
the childhood and working life of the adopted person.1 Unlike, most cost-benefit studies, which are
based upon data generated by program design, this first cost-benefit analysis of adoption must be
based on observations of the adopted and the fostered existing in the clinical and epidemiological
literature.
The conduct of a cost-benefit analysis of adoption from foster care requires several steps:
•
estimate the effects of adoption on the adopted child,
•
estimate the streams of monetary values of benefits and costs associated with the effects,
•
calculate the present value of benefits net of costs, and
•
conduct a sensitivity analysis.
1
The reasons why adoption confers benefits on the adopted are thoroughly discussed elsewhere. To
summarize: “It is inconceivable that [legal] insecurity has not influenced the relationship between
foster parents and the children” (Bohman and Sigvardson, 1990). Adoption, by contrast, “offers
higher levels of emotional security, sense of belonging and general well-being” (Triseliotis 2002).
2
Ideally, cost-benefit analysis includes each consequence of the program to be evaluated.
Consequences may be intended effects of the program or unintended effects. The effects of
adoption used in this analysis include cognitive and educational effects, health and mental health
effects, behavioral effects (including effects on crime), and economic effects. Estimates of the
monetary value of the effects of adoption are made using evidence from labor economics and from
other cost-benefit studies.
I calculate the net present value of three types of adoption from foster care. The first type is
early adoption: a child enters care at the median age of three, experiences stable placement, and is
adopted as quickly as possible after termination of parental rights. The second is late adoption: the
child enters care at age eight, experiences stable placement and speedy termination and adoption.
The third is delayed placement: the child enters care at the median age of three, may experience
unstable placement and/or delays in termination, so that adoption does not occur until the child is
eight.
At several junctures in the conduct of the analysis, it is necessary to make simplifying
assumptions. The final step is to test the sensitivity of the results to the assumptions. The following
sections discuss the steps in detail; a summary of the policy implications of the work concludes the
paper.
II. A. Effects of Adoption on the Adopted
An ideal cost-benefit analysis of adoption from foster care would be based upon a long-run,
detailed, longitudinal study beginning with a random sample of all children at birth. Because foster
care and adoption—and especially adoption from foster care—are relatively rare events, the sample
would need to be quite large to capture enough observations for statistical analysis. No such study
3
exists.2 In the absence of direct evidence I estimate the benefits of adoption using the clinical and
epidemiological evidence accumulated by scholars in various disciplines.
The appendix summarizes the results of dozens of studies of adopted and foster children.
The results of the studies are grouped by type of outcome (e.g. cognitive/educational, economic).
Studies that report on several types of outcomes appear in all the relevant places in the appendix.
While the exact degree of reversibility of childhood trauma is still a matter of debate in the
medical literature, it is clear that when children are removed from dysfunctional and unstimulating
or inappropriately stimulating environments, and subsequently adopted by functioning families, they
experience significant catch-up. Catch-up effects have been measured for victims of abuse and
neglect (Kadushin 1967) and for children rescued from orphanages (Rutter et al 1998). The
literature summarized in the appendix contains studies of both children with rocky starts and
children relinquished and adopted as infants.
Table 1 summarizes the direction of the effects measured by the studies included in the
appendix. Adoption improves good outcomes (such as educational attainment and self-sufficiency)
and reduces negative outcomes (such as delinquency and welfare receipt). Mixed evidence exists
regarding the emotional and psychological health of the adopted; it appears that adoptive parents are
more apt to use mental health services (e.g. Zill 1995).
[insert table 1 about here]
Relatively few studies include subjects in every kind of living arrangement. Many studies
compare those fostered to those raised in intact families. Other studies compare the adopted with
those raised by their birth parents. A few studies compare the adopted with those born out of
2
The National Survey of Child and Adolescent Well-being may eventually provide enough evidence,
but more waves will need to be collected first. Existing long-term studies of adoption focus on
stability of placement and psychological adjustment (see Rushton (2004) and Triseliotis (2002) for
reviews). Psychological health is one of many health outcomes discussed here.
4
wedlock who would have been candidates for relinquishment and adoption in infancy. An even
smaller number of studies identify whether the adopted person was placed in infancy or later in
childhood. Only a tiny subset of all studies of adoption (note especially Bohman and Sigvardsson
1990, Brand and Brinich 1999, Dworsky 2005, and Gibbons et al 1995) allow direct comparison of
persons who were adopted and fostered. These studies, however, alongside the many studies
comparing two groups who were raised in different kinds of families, lead in the same direction.
The preponderance of evidence is that adoption is an effective intervention. Adoption
improves the cognitive, educational, health, social, and economic outcomes relative to the children’s
pre-adoptive status (whether this is a lone parent family as in infant relinquishment or whether it is
foster care). Adopted children do better than the characteristics of their pre-adoptive conditions
would predict; however, they do not do as well as the characteristics of their adoptive conditions
would predict. In other words, catch-up is substantial but not complete.
Table 2 uses this summary fact—that the adopted do better than those fostered-long term
but less well than those in intact two-parent birth families—to arrive at a quantitative estimate of the
effect of adoption on each type of outcome.
[insert table 2 about here]
The cognitive and educational outcomes studied include IQ, educational progress, special
education placements, secondary school completion, and tertiary school attendance. A recent metaanalysis estimates that adoption about doubles measured IQ relative to peers who remain in their
original inadequate environment (van Ijzendoorn et al 2005). Educational progress, measured in
terms of academic ability and performance on grade level, is 50 percent better. Adopted children are
referred to special education about half as often. They are 23 percent more likely to complete high
school or its equivalent, and they are twice as likely to obtain additional schooling after secondary
school.
5
Adoption improves health outcomes, but the magnitude of the health effect is smaller than
the educational effect. The adopted are about four percent more likely to report being healthy. Selfreports are confirmed by evidence from childhood assessments and hospital or emergency room
visits. The adopted are about 20 percent less likely to become parents as teens and about 15 percent
less likely to use or abuse alcohol or other substances.
Much of the adoption literature is based on models of adoption that predict adverse
psychological effects in adolescence because of identity confusion. Thus there is a large literature on
the psychological health of the adopted. The adopted are 200 percent more likely to receive mental
health services, but this does not necessarily indicate that adoption causes psychological problems
(see reviews by Rushton (2004) and Triseliotis (2002)). It may instead indicate that adoptive parents
demand more mental health services than other parents, perhaps because of the fact of adoption,
but also because they are primed to expect problems. There are available measurements of malaise,
depression and overall well-being in the literature. These reports are mixed. In table 2, the estimate
that the adopted are 26 percent less likely to score in the clinical range is based upon the application
of the stylized fact that the adopted do better than those fostered long-term but not as well as the
population in general to the scores reported in the detailed epidemiological study using the 1970
British Birth Cohort Study (Viner and Taylor 2005).
The effect of adoption on externalizing behaviors among the adopted as children and
adolescents is similarly difficult to estimate because the results are mixed. It appears, though, that
differences may be driven by small numbers, so that the median adopted child is unlikely to exhibit
more negative externalizing behaviors than the median child in the population (Brand and Brinich
1999). I do not report an effect for externalizing behaviors in table 2 because these behaviors are
likely to be reflected in better-measured indicators such as school suspension and delinquency.
6
Whatever difficulties the fact of adoption brings to the adopted, objective measures of the
outcomes of adoption are positive. The adopted are 21 percent less likely to be suspended or
expelled from school, are about half as likely to be delinquent or arrested, and are 32 percent less
likely to be incarcerated. A few studies have attempted to measure the relationship skills acquired by
foster children and their overall levels of social connectedness. I estimate that the adopted are 34
percent less likely to have poor interpersonal relationship skills and that they are 60 percent less
likely to be socially disconnected as young adults.
Finally, the adopted are more self-sufficient as young adults than the fostered. The final
category of outcomes reported by the studies summarized in the appendix is the direct measure of
economic outcomes. The adopted are 22 percent more likely to be in the labor force as young
adults; they are 15 percent more likely to be employed; and they are 24 percent less likely to be
unemployed. They have higher incomes (after adjusting for time spent in school) and lower
incidence of welfare program participation.
II. B. Monetary Valuation of Benefits of Adoption
The linchpin of any cost-benefit analysis is the monetary valuation of the benefits. Monetary
valuation is necessary because it is the only way the various effects can be aggregated and compared
to costs. I begin with estimation of the net private value of adoption, or the value to adoption that
is realized by the adopted net of costs to their adoptive families. I then evaluate the net savings to
government and to the public at large. Government savings fall into three main categories: (1)
reduced costs of education, (2) child welfare/human service savings and (3) savings from reduced
crime.
Economists have long recognized the importance of education in adding to earnings power;
recent advances in labor economics and micro-econometrics make it possible to assign monetary
7
values to the other outcomes of adoption in terms of their effects on income. The first column of
table 3 shows the best available estimate of the increment to income associated with each measured
outcome of adoption.3
A one standard deviation increase in IQ is associated with a four percent increase in earning
power, while academic ability is associated with an eight percent increase. Completion of high
school has the largest effect—a 37 percent increase in income. Additional years of education add 10
percent to earning power. Good health adds almost five percent to earning power. Teen
parenthood is valued through its effect on the high school dropout rate. Avoidance of teen
parenthood adds 1.4 percent to income.
A recent theme in the literature on human capital policy has emphasized the importance of
psychological and social functioning on wages (Carneiro and Heckman 2003, Heckman and Lochner
2000). Psychological health, measured in the economics literature as the effect of self-esteem, is the
second-largest contributor to earning power, adding almost one-third. Avoiding suspension or
expulsion increases income by 10.4 percent, while avoiding incarceration increases earning power by
16 percent. Relationship skills, proxied by the effect of sociality in youth, add two percent. Finally,
each additional year of employment experience adds 14 percent to earnings power.
3
It would have been possible to use a large number of studies from economics to estimate the
average effects of some characteristics (such as education and experience) on income. This method
would have resulted in a higher estimate of the effect of these oft-measured characteristics. Instead,
I chose a more conservative route and used only the few, recent studies that isolate the effect of
education and experience from the effect of other characteristics, such as mental health, affected by
adoption. This reduces the chances of double-counting benefits. To simplify the work, I have
omitted the marriage premium for men and marriage penalty for women, assuming they cancel each
other out in the adopted population. I have also omitted race and gender differences in incomes.
8
II. B. i. Private Benefits and Costs
The next step is to aggregate the increments in earnings power. Since earnings are correlated
with age, I estimate the added earnings from adoption at different stages of life. I assume that the
ratio of the earnings of the fostered to the earnings of the median income-earner in the population is
constant across all life stages. While this might not be true, the limited follow-up in existing
longitudinal studies does not allow more precise estimation of the ratio. In fact, however, the
assumption likely results in an underestimate because adoption is associated with greater human
capital accumulation and the wages of the skilled and educated have risen more rapidly in recent
decades than the wages of the unskilled (Carneiro and Heckman 2003, Krueger 2003, among many).
To estimate the incremental monetary benefits of adoption to the adopted at each age, I take
the product of three values: (1) the outcomes of adoption from table 2, (2) the increments to income
from column 1 of table 3, and (3) the estimated earnings by the fostered at each age. The estimated
age-income profiles of the long-term fostered, the adopted, and the median income-earner in the
population are pictured in figure 1. The estimated private benefit to adoption is the sum, across all
ages, of the difference in earnings between the fostered and the adoption. This is the area between
the income profile of the fostered and the income profile of the adopted.
[insert figure 1 about here]
The added income from adoption at age 30 is shown in column 2 of table 3. For the 30year-old who was adopted, income is estimated to be about $9,258 greater than it would have been if
the person had been fostered long-term but never adopted. Note that this total is lower than the
single direct observation of about $16,000 (in 2000 dollars) in additional income inferred from Amy
Dworsky’s (2005) study of young adults in Wisconsin.
[insert table 3 about here]
9
A final, and substantial, private benefit recorded in table 3 is the value to the adopted of
avoiding prison. Based on the work of Joel Waldfogel (1994), I estimate the lifetime private benefit
of avoiding prison to be over a year’s income.
Costs offset some of the private benefits of adoption. I use the USDA (2000) estimates of
average expenditures on children to estimate the added cost to families of helping their adopted
children catch up. I assume that family child care, education, and health expenditures on the
adopted are 25 percent higher than average. Even though many adoptive families are supported by
adoption assistance payments (Dalberth et al 2005, Hansen 2006), adoptive parents report significant
expenditures to help their adopted children compensate for their adverse early experiences (Sedlack
and Broadhurst 1993, Children’s Rights 2006). Additional costs may also be incurred for tertiary
schooling. These I have valued at the current tuition at public two-year colleges.
I do not account for opportunity cost of time lost at work to care for an adopted child.
Implicitly this assumes that the opportunity cost is not different from the opportunity cost of caring
for another child. This may not be the case; however, lost income would have to exceed 3.3 years of
median income to erase the benefits of additional income to the child.
II. B. ii. Public Benefits and Costs
The government savings from adoption in child welfare and human services costs reported
by Richard Barth and his co-authors (2006) is used here; I adjust their estimates only for inflation.
To the human service savings I add the savings to the government and the public from reduced
crime. I assume the crime savings from adoption equal the crime savings from the Perry Preschool
Project (Schweinhart et al 1993), again adjusted for inflation. Since the median age of entry into care
is 3.8 (Barth et al 2006), this seems a reasonable approach. Crime savings include reduced direct
cost and imputed cost of crime to the victims. Because the crime savings were estimated at a time
10
when incarceration costs were lower, the sum of these savings probably underestimates savings
today.
Additional governmental savings from adoption come from reduced grade repetitions and
reduced use of special education services. Additional governmental costs of completing high school
and attending college offset a small fraction of government and public savings.
II. C. Net Present Value
The monetary values in table 3 are all reported in constant 2000 dollars; that is, they are
adjusted for inflation. It is necessary also to adjust for the value of time, since a dollar in hand can
be used for alternatives that also have value. The discount rate, which represents the annual value of
time, is assumed to be 3 percent.
The net present value of the streams of benefits and costs is reported in table 4. The first
column assumes adoptive placement occurs at age 3. This is equivalent to assuming that a young
child with little chance of reunification is placed with foster parents who are interested in adopting.
Such fost-adopt programs are common (Wilson et al 2005). I view this as a best-case scenario:
children who are exposed to adverse conditions for a relatively limited time experience the most
catch up. The net present value of government and public savings from early adoption is nearly
$235,000. The net present value of private benefits is $149,000, which is offset by about $10,000 of
private cost. The net present value of early adoption is thus estimated to be about $375,000.
[insert table 4 about here]
The second column assumes the child enters into care at age 8 and is placed in a stable,
eventually adoptive, family. Human service savings are greater for late adoption than for early
adoption (Barth et al 2006), but I assume that the benefits of late adoption are half of the benefits of
early adoption. The net present value of late adoption is nearly $302,000.
11
The final column assumes the child enters care at age 3 but adoption is delayed until age 8.
Human service savings are lower, and I again assume benefits are half of those experienced by
children adopted early. The net present value of delayed adoption is over $281,000.
Again, an adoption from foster care costs government about $115,000 (Barth et al 2006,
adjusted for inflation to 2000 dollars). Each dollar spent on adoption returns from $2.45 to $3.26 in
benefits to society.
II. D. Sensitivity Analysis
In this section I reconsider key assumptions to check that assumptions do not unduly drive
the findings reported in the previous section. I alter three assumptions; the results are shown in
table 5.
First, I consider the possibility that the discount rate is 6 percent rather than 3 percent. This
reduces the net present value of adoption to between $254,500 and $311,500, depending upon the
age at adoption. Second, I consider that the estimates in table 2 may overstate the true benefits of
adoption. If the benefits estimated in table 2 are halved, the present value of adoption is between
$235,500 and $302,000. If the contributions of outcomes to income reported in table 3 are similarly
overstated, then the net present value may be between $230,000 and $291,000.
Changing the latter two assumptions moves the estimated age-income profile of the adopted
closer to the profile of the fostered in figure 1. This is the same adjustment that would need to be
made if the adopted are a non-random sample of the fostered (that is, if the adopted are positively
selected such that they are the most able to benefit from adoption). Even if all three assumptions
are altered, as shown in the last line of table 5, adoption from foster care has net present value of
between about $214,000 and $259,000.
[insert table 5 about here]
12
III. Limitations of the Study
The absence of a large-scale, long-term longitudinal study of children growing up in different
arrangements limits our ability to estimate the benefits of adoption with precision. The necessary
imprecision of this study, however, is tempered by its self-conscious efforts to underestimate
benefits and overestimate costs.
An advantage having a longitudinal study would be the ability to capture the effect of
selection bias. Selection bias may occur if the youngest and healthiest waiting children are
overrepresented in the adopted population relative to the foster care population from which they are
drawn. Countervailing forces exist in adoption policy, though, because adoption promotion is
targeted towards children with special needs and especially deprived birth family environments. It is
difficult to guess, therefore, the degree to which the benefits of adoption may be overstated in the
existing literature. The sensitivity analysis shows that even if there is significant overstatement of the
benefits of adoption, the results of the study cannot be invalidated.
Finally, the stability of adoptive placement is not addressed here. If a significant number of
the adopted do not remain in their adoptive families, benefits are overstated and child welfare costs
are understated. However, recent reviews (e.g. Rushton 2004) show that adoptive placements, even
of difficult children, are unlikely to break down. Even when children need more restrictive care than
can be provided in the family home, adoptive families maintain responsibility for and contact with
their troubled children.
IV. Policy Recommendations
The benefits of adoption from foster care dwarf the costs, however substantial the costs may
be. If we wish to capture these benefits to society and wish to improve the life chances of our most
13
vulnerable citizens, we must “take adoption seriously” (Bartholet 2000). To take adoption seriously
we must ensure that all waiting children have a high and equal chance of being adopted.
To take adoption seriously we must educate prospective adoptive parents about the benefits
of adoption from foster care: waits can be shorter, some expenses are covered, an unqualified tax
credit is available in the year of finalization, and ongoing help with medical and other expenses can
be negotiated.
To take adoption seriously we must continue to enforce the Multiethnic Placement Acts.
Recent actions taken against Hamilton County, Ohio, and the state of South Carolina indicate that
racial matching policies are less likely to be tolerated (Bartholet 2006). However, even subtle hints
that transracial placements are inappropriate may reduce the chances of adoption for minority
children (Hansen and Pollack 2006).
Finally, to take adoption seriously we must fully fund the federal adoption assistance
program. The system of child welfare financing currently in place does not ensure that all children
in need of adoptive families have an equal chance of being adopted. Family decisions about
adoption are sensitive to post-adoption financial support (Hansen and Hansen 2006), and state
offers of post-adoption financial support are sensitive to the availability of federal funding (Dalberth
et al 2005). In order for the state to claim federal reimbursement of child welfare and adoption
expenditures on a child, the child must have state-defined special needs and must either qualify for
federal Supplemental Security Income or must have been removed from a family that would have
qualified for benefits under Aid to Families with Dependent Children at the time of removal. Of
course, because AFDC was replaced by the Temporary Assistance for Needy Families, AFDC
eligibility criteria are no longer updated. Therefore, fewer waiting children are likely to be IV-E
eligible in the future. States wishing to secure the benefits of adoption from foster care could
therefore see an untenable increase in their fiscal obligations if there is no change in the federal
14
definition of IV-E eligibility. A 2004 Pew Commission report suggested de-linking adoption
assistance from the out-dated AFDC criteria (Pew Commission, 2004), and Senator Rockefeller and
Representatives Herger and Cardin introduced bills intended to free states from the worry that
federal matching funds will disappear. However, none of these bills emerged from committee.4
Capturing the benefits of adoption in the future requires action now.
4
Most recently, Sen. John D. Rockefeller (D-WV) introduced S. 1539 (A bill to amend part E of title
IV of the Social Security Act to promote the adoption of children with special needs) on July 28,
2005. Rep. Benjamin Cardin (D-MD) introduced H.R. 1534 (Child Protective Services
Improvement Act) on April 1, 2003. Rep. Herger introduced H.R. 4856 (Child Safety, Adoption,
and Family Enhancement Act) on July 19, 2004.
15
References
Barth, Richard, P. (1997). “The Value of Special Needs Adoption.” In Rosemary Avery
(ed.), Adoption Policy and Special Needs Children (New York: Auburn House), 171-204.
Barth, Richard P., Chung Kwon Lee, Judith Wildfire, and Shenyang Guo (2006). A comparison of
the governmental costs of long-term foster care and adoption. Social Service Review, 80(1),
127-158.
Bartholet, Elizabeth (2000). Nobody’s Children (Boston: Beacon Press).
Bartholet, Elizabeth (2006). “Multiethnic Placement Act Enforcement Decisions.” ONLINE.
Available: http://www.law.harvard.edu/faculty/bartholet/mepa.php (last accessed
December 6, 2006).
Benedict, Mary I., Susan Zurvain, and Rebecca Y. Stallings (1996). “Adult Functioning of Children
Who Lived in Kin Versus Non-Relative Family Foster Homes.” Child Welfare 75(5): 529549.
Blome, Wendy Whiting (1997). “What Happens to Foster Kids: Educational Experiences of a
Random Sample of Foster Care Youth and a Matched Group of Non-Foster Care Youth.”
Child and Adolescent Social Work Journal 14(1): 41-54.
Bohman, Michael, and Soren Sigvardsson (1990). “Outcome in Adoption: Lessons from
Longitudinal Studies.” In D.M. Brodzinsky and M.D. Schecher (eds). The Psychology of
Adoption (New York: Oxford University Press): 93-106
Borders, L DiAnne; Lynda K Black; B. Kay Pasley (1998). “Are Adopted Children and Their
Parents at Greater Risk for Negative Outcomes?” Family Relations 47(3): 237-241.
Brand, Ann E., and Paul M. Brinich (1999). “Behavior Problems and Mental Health Contacts in
Adopted, Foster and Nonadopted Children.” Journal of Child Psychology, Psychiatry, and Allied
Fields 40(8): 1221-1229.
16
Broghans, Lex, Bas ter Weel, and Bruce A. Weinberg (2006). “People People: Social Capital and the
Labor-Market Outcome.” IZA Discussion Paper 194. ONLINE. Available:
http://ssrn.com/abstract=670207 (last accessed November 20, 2006).
Carneiro, Pedro and James J. Heckman (2003). “Human Capital Policy.” In Alan B. Krueger and
James J. Heckman (eds.), Inequality in America: What Role for Human Capital Policies (Cambridge,
MA: The MIT Press): 77-240.
Cheung, S.Y., and A. Buchanan (1997). “Malaise Scores in Adulthood of Children and Young
People Who Have Been in Care.” Journal of Child Psychology, Psychiatry and Allied Fields 38: 575580.
Children’s Rights (2006). “Ending the Foster Care Life Sentence: The Critical Need for Adoption
Subsidies.” ONLINE. Available:
http://www.childrensrights.org/pdfs/FINAL%20ADOPTION%20SUBSIDY%20DATA
%20REPORT.doc.pdf (last accessed August 20, 2006).
Cook, R. (1991). “A National Evaluation of Title IV-E Foster Care Independent Living Program
for Youth: Phase 2 Final Report.” Manuscript. Rockville, MD: Westat.
Courtney, Mark E., Irving Piviavin, Andrew Grogan-Kaylor, and Ande Nesmith (2001). “Foster
Youth Transitions to Adulthood: A Longitudinal View of Youth Leaving Care.” Child
Welfare 80(6): 685-717.
Courtney, Mark E., Amy Dworsky, Gretchen Ruth, Tom Keller, Judy Havlicek, and Noel Bost
(2005). “Midwest Evaluation of the Adult Functioning of Former Foster Youth.” Chapin
Hall Working Paper. ONLINE. Available:
http://www.chapinhall.org/article_abstract.aspx?ar=1355 (last accessed November 15,
2006).
17
Dalberth, Barbara, Deborah Gibbs, and Nancy Berkman (2005). “Understanding Adoption
Subsidies: An Analysis of AFCARS Data: Final Report.” ONLINE. Available:
http://aspe.hhs.gov/hsp/05/adoption-subsidies/index.htm (last accessed November 30,
2006).
Dumaret, Annick, Camille Coppel-Batch, and Simone Couraud (1997). “Adult Outcomes of
Children Reared for Long-Term Periods in Foster Care.” Child Abuse and Neglect 21(10): 911927.
Dworsky, Amy, and Mark E. Courtney, Mark E. (2000). “Self-Sufficiency of Former Foster Youth
in Wisconsin: Analysis of Unemployment Insurance Wage Data and Public Assistance
Data.” Institute for Research on Poverty Special Report 81. ONLINE. Available:
http://www.irp.wisc.edu/publications/sr/pdfs/sr81.pdf (last accessed December 6, 2006).
Dworsky, Amy (2005). “The Economic Self-Sufficiency of Wisconsin’s Former Foster Youth.”
Children and Youth Services Review 27: 1085-1118.
Dworsky, Amy (2006). Personal communication via e-mail. October 26, 2006.
Eckholm, Erik (2005). “Law Cutting Adoption Payments Is Faulted.” New York Times, Aug. 16, A9.
Fergusson, David M., and L. John Horwood (1998). “Adoption and Adjustment in Adolescence.”
Adoption and Fostering 22(1): 7-15.
Fergusson, David M., Michael Lynskey, and L. John Horwood (1995). “The Adolescent Outcomes
of Adoption: A 16-Year Longitudinal Study.” Journal of Child Psychology, Psychiatry and Allied
Fields 36(4): 597-615.
Festinger, Trudy (1983). No One Ever Asked Us (New York: Columbia University Press).
Gibbons, Jane, Bernard Gallagher, Caroline Bell, and David Gordon (1995). Development after Physical
Abuse in Early Childhood. London: HMSO.
18
Goerge, Robert M. Lucy Bilaver, Bong Joo Lee, Barbara Needell, Alan Brookhart, William Jackman
(2002). “Employment Outcomes for Youth Aging Out of Foster Care.” ONLINE.
Available: http://aspe.hhs.gov/hsp/fostercare-agingout02/ (last accessed December 5,
2006).
Goldsmith, Arthur H., Jonathan R. Veum, and William Darity, Jr. (1997). “The Impact of
Psychological and Human Capital on Wages.” Economic Inquiry 35(October): 815-29.
Mary Eschelbach Hansen (2006). “Title IV-E Claims and Adoption Assistance Payments.”
AFCARS Adoption Data Research Brief Number 5. ONLINE. North American Council
on Adoptable Children. Available:
http://www.nacac.org/pdfs/AFCARStitleivepayments.pdf (last accessed August 24, 2006).
Hansen, Mary Eschelbach, and Bradley A. Hansen (2006). “An Economic Analysis of the Adoption
of Children from Foster Care” (with Bradley Hansen), Child Welfare 85, 3(May/June 2006),
pp. 559-583.
Hansen, Mary Eschelbach, and Daniel Pollack (2006). “The subtleties of race and recruitment in
foster care and adoption.” California NASW News 32, 8: 6.
Hansen, Robin L., Fatema Lakhani Mawjee, Keith Barton, Mary B. Metcalf, and Nancy R. Joye
(2004). “Comparing the Health Status of Low-Income Children In and Out of Foster Care.”
Child Welfare 83(4): 367-380.
Heckman, James J., and Lance Lochner (2000). “Rethinking Education and Training Policy:
Understanding the Sources of Skill Formation in a Modern Economy”. In Sheldon
Danziger and Jane Waldfogel (eds.), Securing the Future (New York: Russell Sage Foundation):
47-83.
Illinois Dept. of Human Services (2006). “Facts about TANF.” ONLINE.
http://www.dhs.state.il.us/ts/fss/tanfStatistics.asp (last accessed November 15, 2006).
19
Jones, Melanie K., Paul L. Latreille, and Peter J. Sloane (2006). “Disability, Gender, and the
British Labour Market.” Oxford Economic Papers 58: 407-449.
Kadushin, A. (1967). “Reversibility of Trauma: A Follow Up Study of Children Adopted
When Older.” Social Work 12: 22-33.
Thomas Kane and Cecelia Rouse (1995). "Labor Market Returns to Two- and Four-year College,"
American Economic Review, 85 no. 3 (June 1995): 600-614.
Kerman, Benjamin, Judith Wildfire, and Richard Barth (2002). “Outcomes for Young Adults Who
Experienced Foster Care.” Children and Youth Services Review 24(5): 319-344.
Kreider, Rose (2003). “Adopted Children and Stepchildren: 2000.” Census 2000 Special Reports.
ONLINE. Available: http://www.census.gov/prod/2003pubs/censr-6.pdf (last accessed 18
May 2006).
Krueger, Alan, B. (2003). “Inequality: Too Much of a Good Thing.” In Alan B. Krueger and James
J. Heckman (eds.), Inequality in America: What Role for Human Capital Policies (Cambridge, MA:
The MIT Press): 1-76.
Levine, David I., and Gary Painter (2003). “The Schooling Costs of Teenage Out-of-Wedlock
Childbearing: Analysis with a Within-School Propensity-Score-Matching Estimator.” The
Review of Economics and Statistics 85(4): 884-900.
MDRC (2006). “Food Stamp Caseload Dynamics.” ONLINE. Available:
http://www.mdrc.org/publications/420/overview.html (last accessed November 15, 2006).
Maughn, Barbara, and Andrew Pickles (1990). “Adopted and Illegitimate Children Growing Up.”
In Lee N. Robbins and Michael Rutter (eds). Straight and Devious Pathways from Childhood to
Adulthood (Cambridge: Cambridge University Press): 36-61.
20
Maughan, Barbara, Stephan Collishaw, and Andrew Pickles (1998). “School Achievement and
Adult Qualifications among Adoptees: A Longitudinal Study.” Journal of Child Psychology,
Psychiatry and Allied Fields 39(5): 669-685.
McCann, J.B., A. James, S. Wilson, and G. Dunn (1996). “Prevalence of Psychiatric Disorders in
Young People in the Care System.” British Medical Journal 313: 1524-1527.
McMillan, J.Curtis, and Jayne Tucker (1999). “The Status of Older Adolescents at Exit from Outof-Home Care.” Child Welfare 78(3): 339-360.
National Law Center on Homelessness and Poverty (2004). “Key Data Concerning Homeless
Persons in America.” ONLINE. Available:
http://www.nlchp.org/FA_HAPIA/HomelessPersonsinAmerica.pdf (last accessed
November 30, 2006).
National Education Association [NEA] (2006). “Special Education and the Individuals with
Disabilities Education Act.” ONLINE. Available:
http://www.nea.org/specialed/index.html (last accessed November 30, 2006).
North American Council on Adoptable Children [NACAC] (2003). “Preserving Adoption Support
Programs: Parents Can Make a Difference. ONLINE.
http://www.nacac.org/subsidyfactsheets/difference.html (last accessed August 23, 2006).
Pew Commission on Children and Foster Care (2004). “Fostering the Future: Safety, Permanence
and Well-Being for Children in Foster Care.” ONLINE. Available:
http://pewfostercare.org/research/docs/FinalReport.pdf (last accessed April 30, 2006).
Plug, Erik and Wim Vijverberg (2003). “Schooling, Family Background, and Adoption: Is It Nature
or Is It Nurture?” Journal of Political Economy 111(3): 611-641.
Prosser, William R. (1997). “Family Structure, Substitute Care, and Educational Achievement.”
Institute for Research on Poverty Discussion Paper 1140-97. ONLINE. Available: http:
21
http://www.irp.wisc.edu/publications/dps/pdfs/dp114097.pdf (last accessed December 5,
2006).
Reilly, Thomas (2003). “Transition from Care: Status and Outcomes of Youth Who Age Out of
Foster Care.” Child Welfare (6): 727-746.
Ringel, Jeanne S., Phillis L. Ellickson, and Rebecca L. Collins (2006). “The Relationship between
High School Marijuana Use and Annual Earnings among Young Adult Males.” Contemporary
Economic Policy 24(1): 52-63.
Rushton, Alan (2004). “A Scoping and Scanning Review of Research on the Adoption of Children
Placed from Public Care.” Clinical Child Psychology and Psychiatry 9(1): 89-106.
Rutter, Michael, and the English and Romanian Adoptees Study Team (1998). “Developmental
Catch-up, and Deficit, Following Adoption after Sever Global Early Privation.” Journal of
Child Psychology, Psychiatry and Allied Disciplines 39(4): 465-476.
Scarcella, Cynthia Andrews; Roseana Bess, Erica Hecht Zielewski, Lindsay Warner, and Rob Geen
(2004). The Cost of Protecting Vulnerable Children IV. (Washington, DC: Urban Institute).
Schweinhart, Lawrence J., Helen V. Barnes, David P. Weikart (1993). Significant Benefits: The
High/Scope Perry Preschool Study through Age 27 (Yipsalanti: The High/Scope Press).
Sedlack, Andrea, and Diane Broadhurst (1993). Study of Adoption Assistance Impact and
Outcomes: Final Report. Manscript. Submitted to the Administration for Children,
Youth and Families, DHHS, Contract No. 105-89-1607.
Taussig, H.N., R.B. Clyman, and J. Landsverk (2001). “Children Who Returned Home from Foster
Care: A 6-year Prospective Study of Behavioral Health Outcomes in Adolescence.” Pediatrics
108: 1-7.
Triseliotis, John, and J. Russell (1984). Hard to place: The outcome of adoption and residential care (London:
Heinemann Educational Books).
22
Triseliotis, John (2002). “Long-Term Foster Care or Adoption? The Evidence Examined.” Child
and Family Social Work. 7: 23-33.
US Bureau of the Census (2006). “Historical Income Tables-Households.” ONLINE. Available:
http://www.census.gov/hhes/income/histinc/h10.html (last accessed November 30, 2006).
USDA (2000). “Expenditures on Children by Families: 2000 Annual Report” ONLINIE.
Available http://www.usda.gov/cnpp/Crc/Crc2000.pdf (last accessed November 30, 2006).
USDA (2006). “Food Stamp Program.” ONLINE. Available:
http://www.ers.usda.gov/Briefing/FoodStamps/ (last accessed November 30, 2006).
US Dept. of Education [DOE] (2006). “Community College Facts at a Glance.” ONLINE.
Available: http://www.ed.gov/about/offices/list/ovae/pi/cclo/ccfacts.html (last accessed
November 30, 2006).
US DOJ (2006). “Prison Statistics.” ONLINE. Available:
http://www.ojp.usdoj.gov/bjs/prisons.htm (last accessed November 30, 2006).
Van Ijzendoorn, Marinus H., Femmie Juffer, and Caroline W. Klein Poelhuis (2005). “Adoption and
Cognitive Development: A Meta-Analytic Comparison of Adopted and Nonadopted
Children’s IQ and School Performance.” Psychological Bulletin 131(2): 301-316.
Viner, Russell M., and Brent Taylor (2005). “Adult Health and Social Outcomes of Children Who
Have Been in Public Care: Population-Based Study.” Pediatrics 115(4): 894-899.
Waldfogel, Joel (1994). “The Effect of Criminal Conviction on Income and the Trust ‘Reposed in
the Workmen’.” The Journal of Human Resources 29(1): 62-81.
Wilson, Julie, Jeff Katz and Rob Geen (2005). “Listening to Parents: Overcoming Barriers to the
Adoption of Children from Foster Care.” KSG Working Paper No. RWP05-005. ONLINE.
Available: http://ssrn.com/abstract=663944 (last accessed September 15, 2006).
23
Wulczyn, Fred, and Kristin Hislop (2002). “Growth in the Adoption Population.” ONLINE.
Available: http://aspe.hhs.gov/HSP/fostercare-issues02/adoption/index.htm (last accessed
November 30, 2006).
Zax, Jeffrey S., and Daniel I. Rees (2002). “IQ, Academic Performance, Environment and
Earnings.” The Review of Economics and Statistics 84(4): 600-16.
Zill, Nicholas (I995). “Adopted Children in the United States: A Profile Based on a National
Survey of Child Health.” Serial 104-33. Testimony before Committee on Ways and Means,
Subcommittee on Human Resources (May 10).
24
Table 1: Summary of Outcomes by Placement Type
Adopted:
Fostered
Adopted:
Kin or Lone
Parent
Adopted:
Population
Foster:
Population
No. of
Studies
Cognitive/Educational Outcomes
Preschool development
+
IQ
Ability/on grade level
+
+
2
+
nd
48
+
-
-
62
+
+
9
Special education referral/placement
Completed high school or equivalent
+
+
+
-
10
College attendance/degree/diploma
+
+
-
-
6
Years of education
+
-
2
-
4
Health Outcomes
Self-reported healthy
+
Prevalence of disability
+
Childhood assessments
+
-
2
Childhood hospital/ER visits
-
+
2
Teen parenthood
+
6
Substance abuse
+
4
-
12
Emotional/psychological health
Overall well-being
mixed
+
mixed
1
+
2
Non-Cognitive or Social Outcome
Behavior problem (externalizing)
-
-
+
+
7
25
nd
+
2
+
+
9
Ever incarcerated
+
2
Homeless
+
3
Suspended/expelled from school
Delinquent or arrested
Relationship skills
-
+
2
-
3
+
-
6
+
-
4
Social connectedness
Economic Outcomes
Labor force participation
Employment
+
Unemployment
-
Income
Welfare participation
-
Self-sufficiency
+
2
+
-
7
-
+
4
2
Notes: nd indicates no difference.
Sources: See Appendix.
26
Table 2: Benefits of Adoption
Estimated Effect
Source/Method:
Cognitive/ Educational Outcomes
IQ
117
percent higher
Meta analysis, van Ijzendoorn et al 2005
Ability/on grade level
50
percent more likely
Meta analysis, van Ijzendoorn et al 2005
Special education
referral/placement
55
percent less likely
Meta analysis, van Ijzendoorn et al 2005
Completed high school or
equivalent
23
percent more likely
Midpoint between foster care and
control/comparison
110
percent more likely
Midpoint between foster care and
control/comparison
Midpoint between foster care and
control/comparison
College attendance/degree/
diploma
Health Outcomes
Self-reported healthy
4
percent more likely to
report healthy
Childhood assessments
7
percent higher on index
Midpoint between single parent and
population
Midpoint between single parent and
population, Zill 1995
Childhood hospital/ER visits
25
percent fewer
Teen parenthood
19
percent less likely
Midpoint between foster care and
control/comparison
Substance abuse
16
percent less likely
Midpoint between foster care and
control/comparison
Emotional/psychological
health
26
percent less likely to be in
clinical range
Midpoint between foster care and
control/comparison, Viner & Taylor 1995
200
percent more likely to seek
treatment
Difference between single parent and adopted,
Zill 1995
Not estimated
Mixed results, correlated with other outcomes
Non-Cognitive or Social Outcomes
Behavior problem
(externalizing)
Suspended/expelled from
school
21
percent less likely
Average of midpoint between foster and pop.
& adopted and pop.
Delinquent or arrested
54
percent less likely
Midpoint between foster care and
27
control/comparison
Ever incarcerated
32
percent less likely
Midpoint between foster care and
control/comparison (a)
Poor relationship skills
34
percent less likely
Midpoint between foster care and
control/comparison
Socially disonnected
60
percent less likely
Midpoint between foster care and
control/comparison
Labor force participation
22
percent more likely
Midpoint between foster care and
control/comparison
Employment
15
percent more likely
Midpoint between foster care and
control/comparison
Unemployment
24
percent less likely
Midpoint between foster care and
control/comparison
Income
75
percent higher
Midpoint between foster care and
control/comparsion (c )
Homeless
47
percent less likely
Midpoint between foster care and
control/comparison (b)
Quarters of TANF receipt
68
percent fewer
Dworsky 2006
Quarters of food stamp
receipt
53
percent fewer
Dworsky 2006
Economic Outcomes
(a) Based on 1995 national incarceration rate. US DOJ 2006.
(b) Based on 2004 homelessness rate. NLCHP 2004.
( c) Based on 1990 median income of $30,000.: US Bureau of the Census 2006
28
Table 3: Monetary Benefits and Costs of Adoption
Addition to
Annual
Income (%)
Annual Addition
to Pre-Tax
Private Income
at Age 30 (a)
Total Addition
to Private
Savings (Cost)
Total Other
Gov't Savings
(Cost)
Child Welfare
Savings from
Adoption (b)
143,302*
Cognitive/ Educational Outcomes
IQ
4.0 (c)
718
Ability/on grade level
8.0 (e)
614
Special education
referral/placement
(4,141)*(d)
3,776 (f)
(g)
999 (f)
Completed high school
or equivalent
37.0 (h)
1,306
College attendance/
degree/diploma
10.0 (i)
1,688
(1,558) (j)
(6,164) (j)
4.6 (k)
28
(2,899)*(l)
(m)
Teen parenthood
1.4 (n)
41
(m)
Substance abuse
7.5 (o)
184
(m)
33.0 (p)
1,316
(1,737) (f)
Health Outcomes
Self-reported healthy
Childhood assessments
Childhood hospital/ER
visits
Emotional/psychological
health
(q)
(m)
Non-Cognitive or Social Outcomes
Behavior problem
(externalizing)
Suspended/expelled
from school
10.4 (r )
734
16.0 (t)
1,669
2.0 (v)
203
1967 (s)
Delinquent or arrested
Ever incarcerated
Poor relationship skills
37,908 (t)
86,383* (u)
29
Socially disconnected
(g)
Economic Outcomes
Labor force participation
(g)
Employment
14.0 (e)
466
Unemployment
14.0 (e)
746
Income
Ever Homeless
16,646 (x)
14.0 (w)
758
Quarters of TANF
receipt
( g)
2,048 (y)
Quarters of food stamp
receipt
( g)
12 (z)
*Net present value (not annual value).
(a) Annual values based on median personal income in constant 2000 dollars (US Bureau of the Census
2006).
(b) Savings from adoption. Assumed entry into care at median age of 3.8 (Barth et al 2006).
(c) Effect of a one standard deviation difference in IQ (Zax and Rees 2002).
(d) Expenditures on child care and education assumed to be 25 percent higher than average in 2000 (USDA
2000).
(e) Goldsmith et al (1997).
(f) Regular and special education costs from NEA (2006).
(g) Benefit not estimated. Assumed to affect income through other channels.
(h) Difference in wages with and without high school degree (Krueger 2003).
(i) Affect of a two year college degree (Kane and Rouse 1995).
(j) Tuition and public expenditure on public two-year colleges from US DOE (2006).
(k) Difference in log wages between disabled and non-disabled. (Jones et al 2006).
(l) Expenditures on child care and education assumed to be 25 percent higher than average in 2000 (USDA
2000). Includes expenditure on mental health.
(m) Included in additional health expenditures. Medicaid included in Barth et al (2006).
(n) 20 percent more likely to dropout (Levine and Painter 2006).
30
(o) Effect of marijuiana use (Ringel et al 2006).
(p) Effect of self-esteem (Goldsmith et al 1997).
(q) Included in additional health expenditure.
( r) Average of breach of trust effects for reported crimes (Waldfogel 1994).
(s) Cost of suspension/expulsion assumed equal to cost of one year special education (NEA 2006).
(t) Average across crimes reported (Waldfogel 1994).
(u) Crime savings of Perry Preschool Project (Schweinhart et al 1993).
(v) Effect of youth sociability on adult wages (Borghans et al 2006).
(w) Value of a year’s work experience (Goldsmith et al 1997).
(x) Difference in income during first two years after emancipation (Dworsky 2006).
(y) Based on average monthly benefit*5 year limit (Illinois Dept. of Human Services 2006).
(z) Based on average monthly benefit*3/36 month limit (USDA ERS 2006, MDRC 2006).
31
Table 4: Net Present Value of Adoption: Baseline Estimates
3 Year Old (a)
8 Year Old (b)
Delayed (c )
Net Present Value of Government Savings
Child Welfare
143,302
167,581
146,581
Crime
86,383
43,192
43,192
Other
4,833
397
397
234,518
211,169
190,169
(8,747)
(4,526)
(4,526)
23,623
13,693
13,693
125,427
74,883
74,883
149,050
88,576
88,576
374,821
302,418
281,418
Subtotal
Net Present Value of Private Costs
Health & Education
Net Present Value of Private Benefits
Incarceration
Income
Subtotal
Total
Discount rate=3%.
In constant 2000 dollars.
(a) Child adopted at earliest possible time after entering care at median age of 3.
(b) Child adopted at earliest possible time after entering care at age 8. Benefits reduced by half. Costs and benefits
realized 5 years earlier.
(c ) Child enters care at 3 but is not adopted until age 8. Additional direct costs of foster care incurred.
32
Table 5: Net Present Value of Adoption: Sensitivity Analysis
3 Year Old
8 Year Old
Delayed
311,481
275,576
254,576
302,253
256,595
235,595
Contribution of benefit to income
overstated by 50 percent
291,633
260,715
239,715
Discount rate of 6% &
Benefit of adoption overstated
270,583
243,174
222,174
Discount rate of 6% &
Contribution to income overstate
267,921
253,718
232,718
Discount rate of 6% &
Benefit of adoption overstated &
Contribution to income overstated
258,765
234,823
213,823
Discount rate of 6%
Benefit of adoption overstated 50 percent
Constant 2000 dollars.
33
Figure 1
Estimated Earnings and Private Benefit of Adoption
Median Earnings in Population
40,000
Income (constant 2000 dollars)
Estimated Earnings after
Adoption
30,000
Private Benefit of Adoption
Undiscounted Sum=$474, 157
Present Value at Age 3=$125,427
20,000
10,000
Estimated Earnings after LongTerm Foster Care
0
20
30
40
50
60
Age
34
APPENDIX 1: Measured Effects of Adoption
Outcome Measure
Adopted
Foster
Care
Single
Parent
Population/
Comparison
Group
Age at time
of study
Time
Location
Source
Notes
Registered
in 1991
Sample of
Preschool
Childen on
English Child
Protection
Registers
Gibbons et
al 1995
Comparison is reported
abused/neglected but
never in out-of-home
care.
van
Ijsendoorn
et al 2005
Cognitive or
Educational
Outcome
Good developmental
outcome
>25%
<15%
20%
11-15
IQ
higher
lower
meta
analysis
varied
IQ
no diff.
no diff.
meta
analysis
varied
18-20
1995-6
WI
Below grade level in
reading
74%
National Child
Development
Study (UK)
National Child
Development
Study (UK)
National Child
Development
Study (UK)
Elementary school
reading level
average
low
average
7
1991
Elementary school
reading level
average
average
average
7
1991
General academic
ability
average
low
average
11
1991
Grades/academic
achievement
higher
lower
meta
analysis
varied
Grades/academic
achievement
lower
higher
meta
analysis
varied
van
Ijsendoorn
et al 2005
Courtney et
al 2001
Maughan et
al 1998
Maughan et
al 1998
Compariosn is peers
"left behind" (i.e.
unadopted). Includes
Romanian & English
adoption study.
Comparison is peers in
adoptive families (i.e.
siblings).
Males. Comparison is
raised by married birth
parents.
Females. Comparison is
raised by married birth
parents.
Maughan et
al 1998
Comparison is raised by
married birth parents.
van
Ijsendoorn
et al 2005
van
Ijsendoorn
Comparison is peers
"left behind" (i.e.
unadopted).
Comparison is peers
"left behind" (i.e.
35
In lower half of class
56%
Linguistic test scores
5.18
Logic test score
5.24
et al 2005
unadopted)
Zill 1995
Comparison is living
with two birth parents.
Living with
grandparents=31%.
31%
38%
under 17
1998
National Health
Interview Survey
4.52
4.86
5.15
18
born mid1950s
Stockholm,
Sweden
4.44
4.87
5.46
18
born mid1950s
Stockholm,
Sweden
Zill 1995
Bohman &
Sigvardsson
1990
Bohman &
Sigvardsson
1990
Male. Tests at military
enlistment.
Male. Tests at military
enlistment.
Comparison is living
with two birth parents.
Living with
grandparents=59%.
Adolescents adopted in
first year of life.
Repeated a grade
16%
60%
13%
under 17
1998
National Health
Interview Survey
Repeated a grade
7%
26%
12%
under 17
1998
National Health
Interview Survey
Zill 1995
Repeated a grade
30%
12%
under 17
1998
National Health
Interview Survey
Zill 1995
Adolescents adopted
after first year of life.
84%
about 16
1980 base
year
High School &
Beyond
Blome 1997
Comparison is national
sample.
17+
1992-93
MO
McMillan &
Tucker 1999
Satisfactory high
school progress
63%
Satisfactory high
school progress
64%
Spatial test scores
5.53
4.77
5.34
5.30
18
born mid1950s
Stockholm,
Sweden
Technical test scores
4.96
4.22
4.40
4.91
18
born mid1950s
Stockholm,
Sweden
18-20
1995-6
WI
meta
analysis
varied
Ever had special
education placement
Special education
referral
37%
+19%
Bohman &
Sigvardsson
1990
Bohman &
Sigvardsson
1990
Courtney et
al 2001
van
Ijsendoorn
et al 2005
Male. Tests at military
enlistment.
Male. Tests at military
enlistment.
Comparison is peers in
adoptive families (i.e.
siblings).
36
Special education
referral
-55%%
meta
analysis
varied
mean of 23
1993-94
Baltimore, MD
van
Ijsendoorn
et al 2005
Benedict et
al 1996
Comparison is peers left
behind (i.e. unadopted).
Graduated high
school/GED
58%
Graduated high
school/GED
77%
93%
about 16
1980 base
year
High School &
Beyond
Blome 1997
Comparison is national
sample.
Graduated high
school/GED
48%
85%
age out
1980s
US National
Cook 1991
Comparison is
population average.
Graduated high
school/GED
63%
18-20
1995-6
WI
Courtney et
al 2001
Graduated high
school/GED
60%
90%
19
2002-04
Midwest
Courtney et
al 2005
Comparison is national
sample.
Graduated high
school/GED
7%
32%
aveage 27.8
left care
1972-84
France, agencybased study
Dumaret et
al 1997
Comparison is "national
norms".
Graduated high
school/GED
69%
mean of 20
late 1990s
NV
Reilly 2003
Left school with no
qualifications
2.90%
17%
9%
33
1991
Left school with no
qualifications
3.40%
24%
14%
33
1991
National Child
Development
Study (UK)
National Child
Development
Study (UK)
1970 British
Birth Cohort
Sutdy
1970 British
Birth Cohort
Sutdy
Maughan et
al 1998
Maughan et
al 1998
Viner &
Taylor 2005
Left school with no
qualifications
40%
29%
30
2000
Left school with no
qualifications
35%
25%
30
2000
Less than high school
62%
52%
22-25
1979
NYC
Festinger
1983
College attendence
9%
18-20
1995-6
WI
Courtney et
al 2001
Viner &
Taylor 2005
Male. Comparison is
raised by married birth
parents.
Female. Comparison is
raised by married birth
parents.
Male. Comparison is
sample of children with
no history of public care.
Female. Comparison is
sample of children with
no history of public care.
Comparison is national
sample.
37
19
2002-2004
Midwest
Courtney et
al 2005
20
late 1990s
NV
Reilly 2003
25%
33
1991
15%
32%
30
2000
26%
33%
30
2000
24%
33
1991
born circa
1939
1957-1992
WI Longitudinal
Survey
Plug &
Vijverberg
2003
mean of 25
1980s
NLSY
Prosser 1997
Single parent group
includes kinship care.
under 17
1998
National Health
Interview Survey
on Child Health
Zill 1995
Comparison is living
with two birth parents.
Living with
grandparents=39%
College attendence
18%
Ever attended college
30%
High attainment (>=2
A levels or
degree/diploma)
High attainment (>=2
A levels or
degree/diploma)
High attainment (>=2
A levels or
degree/diploma)
High attainment (>=2
A levels or
degree/diploma)
24%
10%
28%
Completed years of
education
+1.8
Completed years of
education
+0.4
62%
13%
-1.7
-1.3
National Child
Development
Study (UK)
1970 British
Birth Cohort
Sutdy
1970 British
Birth Cohort
Sutdy
National Child
Development
Study (UK)
Maughan et
al 1998
Viner &
Taylor 2005
Viner &
Taylor 2005
Maughan et
al 1998
Comparison is national
sample.
Male. Comparison is
raised by married birth
parents.
Male. Comparison is
sample of children with
no history of public care.
Female. Comparison is
sample of children with
no history of public care.
Female. Comparison is
raised by married birth
parents.
Comparison is sample
raised by birth parents
Health Outcome
Self-reported
"excellent" health
56%
41
55%
Self-reported healthy
78%
mean of 23
1993-94
Baltimore
Benedict et
al 1996
Self-reported healthy
54%
mean of 20
late 1990s
NV
Reilly 2003
Self-reported healthy
78%
30
2000
1970 British
Birth Cohort
Sutdy
Viner &
Taylor 2005
85%
Male. Comparison is
sample of kids with no
history of public care.
38
Self-reported healthy
79%
85%
30
2000
1970 British
Birth Cohort
Sutdy
Viner &
Taylor 2005
Female. Comparison is
sample of kids with no
history of public care.
At least one disability
12%
5%
5-17
2000
US Census
Kreider
2003
All adopted children.
Mental disability
10%
4%
5-17
2000
US Census
Kreider
2003
All adopted children.
Health status index
(median=50)
51.7
45.4
50.9
under 17
1998
National Health
Interview Survey
Zill 1995
Adolescents adopted in
first year of life.
Health status index
(median=50)
51.5
45.4
50.9
under 17
1998
National Health
Interview Survey
Zill 1995
Adolescents adopted
after first year of life.
Medical/dental
assessments
worse
better
5
1990s
Sacramento, CA
Hansen et al
2004
Comparison is Medicaideligible kids not in care.
ER Visits
more
less
19
2002-2004
Midwest
Courtney et
al 2005
Comparison is national
sample.
Zill 1995
Hospital visits in past
year (per 100
children)
2.1
8.2
Teen fatherhood
13%
Teen motherhood
19%
Teen motherhood
32%
4.5
under 17
1998
National Health
Interview Survey
6%
19
2002-2004
Midwest
Courtney et
al 2005
18-20
1995-6
WI
Courtney et
al 2001
19
2002-2004
Midwest
Courtney et
al 2005
23
born in
1958
UK National
Child
Development
Study
Maughn &
Pickles 1990
12%
Teen motherhood
lower
higher
Teen motherhood
50%
17+
1992-93
MO
McMillan &
Tucker 1999
Teen motherhood
70%
mean of 20
late 1990s
NV
Reilly 2003
Comparison is living
with two birth parents.
Living with
grandparents=39%
Comparison is national
sample.
Comparison is national
sample.
39
30
2000
1970 British
Birth Cohort
Sutdy
Viner &
Taylor 2005
20s
1980s
National
Cook 1991
19
2002-04
Midwest
Courtney et
al 2005
17+
1992-93
MO
McMillan &
Tucker 1999
26%
30
2000
13%
30
2000
13%
mean of 23
1993-94
Baltimore
Benedict et
al 1996
Emotionally disturbed
38%
20s
1980s
National
Cook 1991
Ever received psych.
services
32%
15%
22-25
1979
NYC
Festinger
1983
Comparison is national
sample.
Teen motherhood
3%
Substance
abuse/treatment
50%
Substance
abuse/treatment
7.50%
Substance
abuse/treatment
13%
Used illicit drugs in
last year
34%
Used illicit drugs in
last year
15%
Clinical level of psych.
symptoms
3%
3%
1970 British
Birth Cohort
Sutdy
1970 British
Birth Cohort
Sutdy
Viner &
Taylor 2005
Viner &
Taylor 2005
Comparison is sample of
kids with no history of
public care.
Comparison is national
sample.
Male. Comparison is
sample of kids with no
history of public care.
Female. Comparison is
sample of kids with no
history of public care.
Ever received psych.
services
13%
15%
5%
under 17
1998
National Health
Interview Survey
Zill 1995
Adolescents adopted in
first year of life
Ever received psych.
services
42%
15%
5%
under 17
1998
National Health
Interview Survey
Zill 1995
Adolescents adopted
after first year of life.
17+
1992-93
MO
McMillan &
Tucker 1999
16
born in
1977
New Zealand
Fergusson et
al 1995
mean of 23
1993-94
Baltimore
Benedict et
al 1996
13-17
1980s
UK
McCann et
al 1996
Inpatient psych. care
Internalizing
behaviors/disorders
44%
low
low
Psychiatric diagnosis
18%
Psychiatric diagnosis
67%
15%
Comparison is matched
group with no CPS
contact.
40
Score for malaise
Score for malaise
higher
lower
lower
higher
23&33
born in
1958
23
born in
1958
Score in clinical range
for malaise
20%
13%
30
2000
Score in clinical range
for malaise
29%
19%
30
2000
intermediate
1990-91
Registered
in 1991
while
children
1987-88
Self-reported
depression scale
higher
lower
Overall child wellbeing (index)
no diff.
no diff.
Personal well-being
(composite score)
better
worse
UK
UK National
Child
Development
Study
1970 British
Birth Cohort
Sutdy
1970 British
Birth Cohort
Study
Sample of
Preschool
Childen on
English Child
Protection
Registers
Natl Survey
Families
Households
(US)
Cheung &
Buchanan
1997
Comparison is reported
abused/neglected but
never in out-of-home
care.
Maughn &
Pickles 1990
Viner &
Taylor 2005
Viner &
Taylor 2005
Gibbons et
al 1995
Male. Comparison is
sample of kids with no
history of public care.
Female. Comparison is
sample of kids with no
history of public care.
Comparison is reported
abused/neglected but
never in out-of-home
care.
Borders et al
1998
average of
23
mid 1990s
Casey Family
Services (US)
Kerman et al
2002
Exited foster care
without extended
services. No difference
between extended
services group and
adopted group.
Non-Cognitive or
Social Outcome
Behavior problem
index (median=50)
51.3
54.4
48.4
under 17
1998
National Health
Interview Survey
Zill 1995
Adolescents adopted in
first year of life
Behavior problem
index (median=50)
58.6
54.4
48.4
under 17
1998
National Health
Interview Survey
Zill 1995
Adolescents adopted
after first year of life.
41
7.1
in
childhood/
adolesence
7.1
7.1
10
16.7
7
16.7
Behavior problem
index (of 52)
10
7.1
Behavior problem
index (of 52)
12
Behavior problem
index (of 56)
Behavior problem
index (of 56)
1988
National Health
Interview Survey
Brand &
Brinich 1999
in
childhood/
adolesence
1988
National Health
Interview Survey
Brand &
Brinich 1999
7.4
in
childhood/
adolesence
1988
National Health
Interview Survey
Brand &
Brinich 1999
7.4
in
childhood/
adolesence
1988
National Health
Interview Survey
Brand &
Brinich 1999
19
2002-2004
Midwest
Courtney et
al 2005
average of 5
1998-99
Sacramento CA
Hansen et al
2004
13-17
1980s
UK
McCann et
al 1996
16
born in
1977
New Zealand
7-12
1990-91
San Diego
16
born in
1977
New Zealand
Behavior problems
Behavior problems
42%
Behavior problems
53%
Mean number of
behavior problems
1.25
Behavior problems
Externalizing
behaviors/ disorders
12%
1.83
less
lower
higher
Expelled from school
7%
2%
30
2000
Expelled from school
(Female)
4%
1%
30
2000
5
under 17
1998
Suspended or expelled
from school
6
17
1970 British
Birth Cohort
Sutdy
1970 British
Birth Cohort
Sutdy
National Health
Interview Survey
Age 11-17, adopted
before 6 months.
Compiarson is resides
with birth parents.
Age 11-17, adopted after
6 months. Comparison
is resides with birth
parents.
Age 5-11, adopted
before 6 months.
Comparison is resides
with birth parents.
Age 5-11, adopted after
6 months. Comparison
is resides with birth
parents.
Comparison is national
sample.
Comparison is matched
group with no CPS
contact.
Fergusson &
Horwood
1998
Taussig et al
2001
Fergusson et
al 1995
Viner &
Taylor 2005
Viner &
Taylor 2005
Zill 1995
Male. Comparison is
sample of kids with no
history of public care.
Female. Comparison is
sample of kids with no
history of public care.
Comparison is living
with two birth parents.
42
Living with
grandparents=11%
Appears on criminal
or excise board
register
Arrests
18%
29.20%
16.50%
15.50%
28%
Arrests
22%
Convicted in a court
(criminal or civil)
41%
Convicted in a court
(criminal or civil)
9%
Delinquency
23
born mid1950s
Stockholm,
Sweden
19
2002-2004
Midwest
NYC
Bohman &
Sigvardsson
1990
Courtney et
al 2005
Festinger
1983
Comparison is
population.
Comparison is national
sample.
Foster is discharged
from foster family home;
comparison is discharged
from group facility.
Male. Comparison is
sample of kids with no
history of public care.
Female. Comparison is
sample of kids with no
history of public care.
22-25
1979
23%
30
2000
4%
30
2000
38%
mean of 23
1993-94
Baltimore
Benedict et
al 1996
Delinquency
25%
20s
1980s
National
Cook 1991
Delinquency
18%
18-20
1995-6
WI
Courtney et
al 2001
Delinquency
8%
17+
1992-93
MO
McMillan &
Tucker 1999
Delinquency
45%
mean of 20
late 1990s
NV
Reilly 2003
Delinquency
9%
7-12
1990-91
San Diego
Taussig et al
2001
Pulled a weapon on a
person
8%
3%
19
2002-2004
Midwest
Courtney et
al 2005
Male. Comparison is
national sample.
Pulled a weapon on a
person
4%
0.30%
19
1992-93
Midwest
Courtney et
al 2005
Female. Comparison is
national sample.
1970 British
Birth Cohort
Sutdy
1970 British
Birth Cohort
Sutdy
Viner &
Taylor 2005
Viner &
Taylor 2005
43
Ever incarcerated
18%
18-20
1995-6
WI
Courtney et
al 2001
Ever incarcerated
41%
mean of 20
late 1990s
NV
Reilly 2003
Present incarceration
rate
7%
mean of 23
1993-94
Baltimore
Benedict et
al 1996
Present incarceration
rate
7%
mean of 20
late 1990s
NV
Reilly 2003
Ever homeless
27%
mean of 23
1993-94
Baltimore
Benedict et
al 1996
Ever homeless
12%
18-20
1995-6
WI
Courtney et
al 2001
Ever homeless
36%
mean of 20
late 1990s
NV
Reilly 2003
Ability to form
relationships/ parent
own kids
Ability to form
relationships/parent
own kids
Disconnected (no job,
not in school)
more
able
less able
20s
1970s
Scotland
more
able
less able
mena of 18
born mid1950s
Stockholm,
Sweden
19
2002-2004
Midwest
31%
12%
Triseliotis &
Russell 1984
Bohman &
Sigvardson
1990
Courtney et
al 2005
Disconnected (no job,
not in school)
45%
17+
1993-94
MO
McMillan &
Tucker 1999
Social integration
no diff.
aveage 27.8
1990-92
France, agencybased study
Dumaret et
al 1997
Adopted children placed
at age>= 3. Children
raised in institutions
were less positive than
those in foster care.
Compairson is national
sample.
Relative to normed scale.
Economic
Outcomes
44
Labor force
particpation
48%
20s
1980s
National
Cook 1991
Labor force
particpation
61%
18
1995-6
WI
Courtney et
al 2001
Labor force
particpation
40%
58%
19
2002-2004
Midwest
Courtney et
al 2005
Comparison is national
sample.
Labor force
particpation
less
more
18+
1990s
CA, Il, SC
Goerge et al
2002
Comparison group is
national sample.
Labor force
particpation
38%
17+
1993-94
MO
McMillan &
Tucker 1999
Labor force
particpation
63%
mean of 20
late 1990s
NV
Reilly 2003
Any income from
employment in past
year
78%
90%
19
2002-2004
Midwest
Courtney et
al 2005
Employed
68%
93%
22-25
1979
NYC
Festinger
1983
Employed
53%
74%
22-25
1979
NYC
Festinger
1983
WI
Dworsky
2005,
Dworsky
2006
Quarters employed
during first 2 years
after discharge
Unemployed >6
months of last 5 years
4.1
lower
Unemployed, no
reported occupation
Annual earnings 2
years after discharge
3.7
higher
11%
$2,996
$3,597
21%
$2,215
17 at
discharge
2000
23
born in
1958
average of
27.8
1990-92
17 at
discharge
2000
UK National
Child
Development
Study
France, agencybased study
WI
Comparison is national
sample.
Black males.
Comparison is national
sample.
Black males in NYC.
Comparison is NYC
youth.
Single parent is kinship
care; foster care is
discharged to
independent living.
Maughn &
Pickles 1990
Dumaret et
al 1997
Comparison is
populatioin average.
Dworsky
2005,
Dworsky
2006
Single parent is kinship
care; foster care is
discharged to
independent living.
45
Income from
employment in past
year<$10,000
Median annual
income
90%
$3,011
79%
$3,664
19
2002-2004
Midwest
Courtney et
al 2005
17+ at exit
1990s
WI
Dworsky &
Courtney
2000
18+
1990s
CA, Il, SC
Goerge et al
2002
Median annual
income
worse
Median annual
income
<$10,000
mean of 20
late 1990s
NV
Reilly 2003
Median annual
income
$15,000
mean of 23
1993-94
Baltimore
Benedict et
al 1996
2000
WI
better
Dworsky
2005,
Dworsky
2006
Dworsky
2005,
Dworsky
2006
Courtney et
al 2005
Comparison is national
sample.
Adopted includes all
permanent outcomes.
Foster is aged out to
indepnedent living.
Comparison group is
national sample.
Single parent is kinship
care; foster care is
discharged to
independent living.
Single parent is kinship
care; foster care is
discharged to
independent living.
Comparison is national
sample.
Quarters of food
stamps receipt in 2
years after discharge
1.7
3.4
2.4
17 at
discharge
Quarters of TANF
receipt in 2 years after
discharge
0.5
1
1
17 at
discharge
2000
WI
19
2002-2004
Midwest
18-20
1995-6
WI
Courtney et
al 2001
19
2002-2004
Midwest
Courtney et
al 2005
Comparison is national
sample.
Kerman et al
2002
Exited foster care
without extended
services. No difference
between extended
services group and
adopted group.
Received food stamps
17%
Welfare program
participation
32%
Not enough money to
pay rent
12%
Self-sufficiency
(composite score)
better
worse
3%
6%
average of
23
mid 1990s
Casey Family
Services (US)
46